new

Get trending papers in your email inbox once a day!

Get trending papers in your email inbox!

Daily Papers

byAK and the research community

Nov 9

Submitted by

ritianyu

SplatAnything: Gaussian Spacing Normalization for Feed-forward 3DGS across Arbitrary Scenes

zju

Zhejiang University

Submitted by

Sinuo

P^2-MOPD: Accelerating Multi-Teacher On-Policy Distillation via Parameter Priors

stepfun-ai

Submitted by

Hykaruu

Confucius4-R2T2: A Stable-Prefix Streaming ASR Model for Real Real-Time Transcription

netease-youdao

Submitted by

liqiang888

PCBRouteAgent: How Close Are LLM Agents to Becoming PCB Routing Engineers?

·
5 authors

Submitted by

Changqian

LongBanana: An Expert-Verified Benchmark for Long-Context Multi-Reference Image Synthesis

Kwai-Kolors

Kolors Team, Kuaishou Technology

Submitted by

lfqian

LLMs Don't Know What They Know: Benchmarking Temporal Self-Knowledge for Financial Backtesting

TheFinAI

3

Submitted by

Yofuria

Beneath the Surface: What We Overlook When We Talk About Agent Evolution

ustc-community

University of Science and Technology of China

3

Submitted by

delyanboychev

Fast and Exact Relative-Position Attention via Structure-Aware Kernels

INSAIT-Institute

Institute for Computer Science, Artificial intelligence and Technology

2

Submitted by

Tommy-Bie

EvoWorld: Towards Out-of-Sight World Evolution in Generative Video Models

·
8 authors

Submitted by

ChaoyangWang

Video-Coder: Reinforcing Agentic Coding for Video Generation

University of Central Florida

Submitted by

lockon

AgentBaton: Efficient Model Collaboration in Agentic Tasks

nvidia

4

Submitted by

teresadelgadoheras

LowRAM: Low-Rank Keys and Amortised Values for Composable KV Cache Compression at Test-Time

·
5 authors

Submitted by

taesiri

Frontier3DBench: Benchmarking Coding Agents for 3D Creation across Objects, Scenes, and Interactive Open Worlds

·
13 authors

Submitted by

xioamiyh

Exploring Weight-Space Programming: Capabilities, Generalization, and Limitations

·
8 authors

Submitted by

sheliang

AutoCompress: Automated Hardware-Guided Model Compression

nvidia

3

Submitted by

ckarouzos

Steerable Multi-Objective Group-Relative Preference Optimization with Jointly Trained Low-Rank Adapters

·
2 authors

Submitted by

pbelcak

CPU-ARC: Agentic Reasoning Caching for on-CPU Reasoning Reuse

nvidia

Submitted by

chuxuecao

Authorization Drift in LLM Agents: Evaluating Boundary Compliance in Multi-Step Tool Use

·
10 authors

Submitted by

learner-subodh

Human-Centred, Risk-Constrained Evaluation of Financial Document Recognition: Evidence, Critical Errors, and Selective Automation

·
1 authors

3

Submitted by

thuwzt

Hierarchical Residuals: A Minimalist Decoupling of Residual Width from Compute Width

ByteDance-Seed

Submitted by

ethanning

"PRISM" Is Not All You Need: Cross-Topic Name Collisions and Scientific Paper Retrieval

CarnegieMellonU

Carnegie Mellon University

Submitted by

beomjin-ahn

PREEN: Exemplar-Based Color Style Transfer with Smartphone Editor Presets

SamsungResearch

Samsung Research

2

Submitted by

taesiri

OmniLatent: Unified Latent for 3D Part Modeling

·
9 authors

Submitted by

chrisvoncsefalvay

IWIBench: Evaluating Gradient-Free Model Driven Intra-weight Injection

hcltech-robotics

HCLTech Robotics

Submitted by

jaylacho

Knowing When to Teach: Learning from Teachable Prefixes in Multilingual Reasoning Distillation

·
7 authors

3

Submitted by

hazzac

OneTrans-GSR: A Generative Slate Ranking Paradigm Unifying Reranking and Mix-Ranking with One Autoregressive Transformer

ByteDance

2

Submitted by

AllenJin

The Last Mile to the Real World: Agent-Native Primitives as Reusable Building Blocks for Solving Real-World Tasks

·
14 authors

3

Submitted by

mohsenvand

Sketchpad: certified exact arithmetic with a small vision-language model

·
1 authors

Submitted by

svitrack

Toward the Era of Agentic Embodied Intelligence

Submitted by

farenga

Learning to evolve by matching velocities

PolitecnicodiMilano

Politecnico di Milano

Submitted by

Jiaqi-hkust

Where to Connect? Boosting MLLMs via Dynamic Gated Pathways across ALL ViT and LLM Layers

Submitted by

EdeBruyn

Static Recovery Is Not Dynamic Stability: Dynamics-Aware Benchmarking of Protein Motif Scaffolding

HFMI-PROFOUND

Submitted by

ytgui

Coding or Cheating? An Investigation of Frontier Coding LLMs

logitly-ai

Submitted by

JJindou

ODE Policy: Robot Learning via Action Dynamics, Not Actions

ntumars